<font face="trebuchet ms,sans-serif">Thiago,<br><br>Já foi recomendado o não envio de arquivos anexos nas mensagens. Leia o guia de postagem. Sugestão mais simples para conjuntos pequenos é usar dput() como eu faço abaixo.<br>
O problema observado já foi solucionado pelo Fernando. O modelo em questão é não identificável da forma como foi escrito. Existem dois termos x^2 que o torna não estimável. Outro ponto é que o modelo é linear. Veja o código abaixo.<br>
<br><span style="font-family: courier new,monospace;"># lendo da área de transferência (ctrol+c no arquivo texto)</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">da <- read.table("clipboard", header=TRUE, dec=",")</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">str(da)</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># ao dar o dput() gero a estrutura que copio do console e colo</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;"># aqui para que todos tenham acesso imediato (código mínimo REPRODUZÍVEL)</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">dput(da)</span><br style="font-family: courier new,monospace;">
<br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># esse é o resultado do dput() que agora joguei para objeto</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">da <- structure(list(Arvore = 1:100,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                     DAP = c(8.3, 13.6, 13.4, 20.4,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       10.1, 14.4, 13.1, 15.6, 13.4, 13.6, 14.4, 4.4, 10.4, 10.4, 11.6,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       12.1, 12.4, 13.4, 10.6, 14.6, 12.2, 14.1, 10.4, 7.2, 4.4, 10.1,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       14.6, 8.1, 15.1, 14.1, 11.6, 16.3, 11.4, 13.6, 11.4, 13.1, 13.4,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       13.1, 11.4, 10.1, 12.4, 17.4, 10.3, 10.4, 7.4, 14.6, 15.3, 13.6,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       10.4, 12.1, 10.6, 12.2, 9.4, 10.6, 15.3, 7.1, 16.8, 16.1, 19.6,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       8.4, 12.4, 15.4, 12.6, 12.6, 12.6, 9.4, 10.5, 10.1, 13.4, 8.1,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       11.6, 14.4, 11.8, 13.6, 15.6, 10.3, 8.6, 9.4, 9.1, 12.4, 11.4,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       5.2, 11.6, 12.5, 7.5, 8.4, 9.3, 5.1, 18.4, 16.1, 6, 9.4, 12.5,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       9.5, 11.2, 11.5, 6.5, 11.6, 9.5, 11.5),</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                     HT = c(9, 8.65, 10.75, 11.87, 9.14, 9.95, 9.7, 10.5, 9, 8.55, 9.9,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       5.15, 9.25, 8.78, 8.3, 8.8, 9.65, 10.93, 9.25, 9.48, 10.25, 10,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       9.26, 7.47, 5.83, 8.4, 10.45, 7.67, 11.1, 10.95, 11.1, 11.7, 10.36, 11.15, 8.8,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       0.05, 9.6, 11.5, 10.4, 10.65, 10.77, 10.62, 11.68, 8.45, 5.64,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       11.37, 11.1, 12.47, 9.65, 9.85, 10.3, 9.86, 5.9, 9.92, 9.36,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       8.24, 11.68, 12.13, 12.21, 8.2, 10, 11.16, 10.35, 8.1, 10.3,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       8.87, 8.75, 8.8, 9.45, 8.75, 10.45, 11.82, 10.55, 10.15, 11.1,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       10.23, 8.25, 10.3, 9.7, 9.72, 10.27, 5.9, 9.6, 10.1, 8.12, 8.55,</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                       9.17, 5.47, 12.68, 10.5, 6.88, 9.72, 11.5, 10.18, 10.5, 10.26,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                       7.54, 8.7, 9.7, 8.43)),</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">                .Names = c("Arvore", "DAP", "HT"),</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">                class = "data.frame", row.names = c(NA, -100L))</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">dput(da) # isso é o dput(), use esse mecanismo para fornecer CMR à lista</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;"># gráfico e chutes</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">a <- 6; b <- 0.16; c <- -0.125</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">plot(HT~DAP, da)</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">curve(1.30+(x^2/a+b*x+c*x^2), add=TRUE)</span><br style="font-family: courier new,monospace;">
<br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># esse modelo tem dois termos multiplicando x^2</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># temos c*x^2 e temos (1/a)*x^2, portando, não é possível estimá-los</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;"># um termo é função linear do outro</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># gradiente do modelo não linear avaliados no ponto a, b, c</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">grad <- deriv3(~1.30+(x^2/a+b*x+c*x^2),</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">               c("a","b","c"),</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">               function(a, b, c, x) NULL)</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">grad.eval <- grad(a=a, b=b, c=c, x=da$DAP)</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">str(grad.eval)</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">F <- attr(grad.eval, "gradient")</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">eigen(t(F)%*%F)      # último autovalor é praticamente zero</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">solve(t(F)%*%F)      # sistema é singular</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">plot(F[,c("a","c")]) # linha perfeita diz que a é função de c (ou contrário)</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;"># porém, a+c talvez seja uma função estimável</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">ac <- -.01; b <- 0.8</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">plot(HT~DAP, da)</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">curve(1.30+(ac*x^2+b*x), add=TRUE)</span><br style="font-family: courier new,monospace;">
<br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">n0 <- nls(HT~(ac*DAP^2+b*DAP)+1.3, data=da,</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">          start=list(ac=ac, b=b))</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;">summary(n0)</span><br style="font-family: courier new,monospace;"><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># veja que esse modelo é linear e portanto pode-se usar lm()</span><br style="font-family: courier new,monospace;">
<span style="font-family: courier new,monospace;"># o problema é definir b0=1.3 na lm() que não sei se é possível</span><br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;"># perceba que foi necessário apenas uma interação para convergir</span><br style="font-family: courier new,monospace;">
<br style="font-family: courier new,monospace;"><span style="font-family: courier new,monospace;">#-----------------------------------------------------------------------------</span><br style="font-family: courier new,monospace;">
<br style="font-family: courier new,monospace;">Tenho códigos de ajuste de modelos não lineares à dados de DAP nos meus materiais (<a href="http://www.leg.ufpr.br/~walmes/cursoR/cursoR4.pdf">http://www.leg.ufpr.br/~walmes/cursoR/cursoR4.pdf</a>)<br>
<br>À disposição.<br>Walmes.<br><br clear="all"></font><span style="font-family: trebuchet ms,sans-serif;">==========================================================================</span><br style="font-family: trebuchet ms,sans-serif;">
<span style="font-family: trebuchet ms,sans-serif;">Walmes Marques Zeviani</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">LEG (Laboratório de Estatística e Geoinformação, 25.450418 S, 49.231759 W)</span><br style="font-family: trebuchet ms,sans-serif;">
<span style="font-family: trebuchet ms,sans-serif;">Departamento de Estatística - Universidade Federal do Paraná</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">fone: (+55) 41 3361 3573</span><br style="font-family: trebuchet ms,sans-serif;">
<span style="font-family: trebuchet ms,sans-serif;">VoIP: (3361 3600) 1053 1173</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">e-mail: <a href="mailto:walmes@ufpr.br" target="_blank">walmes@ufpr.br</a></span><br style="font-family: trebuchet ms,sans-serif;">
<span style="font-family: trebuchet ms,sans-serif;">twitter: @walmeszeviani</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">homepage: <a href="http://www.leg.ufpr.br/%7Ewalmes" target="_blank">http://www.leg.ufpr.br/~walmes</a></span><br style="font-family: trebuchet ms,sans-serif;">
<span style="font-family: trebuchet ms,sans-serif;">linux user number: 531218</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">==========================================================================</span><br>

<br><br><div class="gmail_quote">2011/8/9 Fernando Henrique Toledo <span dir="ltr"><<a href="mailto:fernandohtoledo@gmail.com">fernandohtoledo@gmail.com</a>></span><br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<span><span style="background-color: rgb(235, 239, 249);" title=""Your model is over-parametrized: d1*exp(-gt) gives two parameters for one constant. As a result, the least-square surface is flat in one direction, and the gradient matrix is singular. "">Tiago<br>

<br>Meus conhecimentos de MNL são torpes, mas as poucas vezes que os enfrentei esse erro apareceu algumas...<br><br>Essa mensagem é um indicativo de que seu modelo pode estar superparametrizado, aí a superfície dos mínimos quadrados é plana e a matriz tem gradiente singular!<br>

<br>Nesses casos vale a pena "esmiuçar" o modelo... </span></span><br><br>Daqui a pouco o Walmes manda uma solução definitiva e elegante sobre isso!<br><br>abraço,<br>FH<br><br><div class="gmail_quote"><div><div>
</div><div class="h5">2011/8/8 Thiago De paula protásio <span dir="ltr"><<a href="mailto:depaulaprotasio@yahoo.com.br" target="_blank">depaulaprotasio@yahoo.com.br</a>></span><br>
</div></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div><div></div><div class="h5"><table border="0" cellpadding="0" cellspacing="0">
<tbody><tr><td style="font: inherit;" valign="top">
Pessoal;<div><br></div><div>Novamente recorro a ajuda de vcs...</div><div>Estou tendo dificuldades no ajuste de um modelo não linear aos meus dados.</div><div><br></div><div>Vejam mensagem de erro que está aparecendo.</div>

<div><br></div><div><div>> a=6</div><div>> b=0.16</div><div>> c=-0.125</div><div><br></div><div>> plot(HT~DAP,data=dados)</div><div>> curve(1.30+(x^2/a+b*x+c*x^2),add=T)</div><div><br></div><div>> n <- nls(HT~(DAP^2/a+b*DAP+c*DAP^2)+1.3,data=dados,</div>

<div>+           start=list(a=a, b=b, c=c))</div><div>Erro em nlsModel(formula, mf, start, wts) : </div><div>  matriz gradiente singular nas estimativas iniciais dos parâmetros</div><div><br></div><div>Em anexo segue a base de dados!</div>

<div><br></div><div>Muito obrigado pela ajuda!</div><div><br></div><font color="#888888"><br>Thiago de Paula
 Protásio<br>
Acadêmico de Engenharia Florestal<br>
Universidade Federal de Lavras<br>
Laboratório de Energia da Biomassa Florestal<br>
Laboratório de Biomateriais<br>
(035) 9183-2246</font></div></td></tr></tbody></table><br></div></div>_______________________________________________<br>
R-br mailing list<br>
<a href="mailto:R-br@listas.c3sl.ufpr.br" target="_blank">R-br@listas.c3sl.ufpr.br</a><br>
<a href="https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br" target="_blank">https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br</a><br>
Leia o guia de postagem (<a href="http://www.leg.ufpr.br/r-br-guia" target="_blank">http://www.leg.ufpr.br/r-br-guia</a>) e forneça código mínimo reproduzível.<br></blockquote></div><br>
<br>_______________________________________________<br>
R-br mailing list<br>
<a href="mailto:R-br@listas.c3sl.ufpr.br">R-br@listas.c3sl.ufpr.br</a><br>
<a href="https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br" target="_blank">https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br</a><br>
Leia o guia de postagem (<a href="http://www.leg.ufpr.br/r-br-guia" target="_blank">http://www.leg.ufpr.br/r-br-guia</a>) e forneça código mínimo reproduzível.<br></blockquote></div><br>